Lennart: The State & Future Of Systemd

[gens]

I just want to add one consideration: an init system + demon management that doesn't use all the linux kernel's features and capability for the sake of compatibility with other kernel is, to me, a giant mountain of shit that should immediately burn with fire.

you would not believe how much the kernel offers that is not used by many programs
amongst those things are a couple that would help something like the systemd "manager" to work more efficiently and without other dependencies

the current init systems already use whatever the kernel offers, even shell based ones
also there are what you could call "service managers" that do take advantage of what the kernel offers to manage processes (daemon is a process, ofc)

[gens]

lots of words.

no need to post another link, just read the interview you yourself quoted
slackware won't switch unless PV is forced to and that, if you ask me, won't happen
also that interview was before eudev, afaik

as for gnome and kde
gnome is the only one now that is dependent
maybe wait for kde to play out then say something


in the end all that functionality is provided by the kernel (and what isn't is basic logic) so anything can be replaced (even console-kit/logind)
nothing in the kernel will ever depend on anything systemd specific, as kernel devs usually have a brain

PS please don't insult people that do not agree with your view of the world

[interested]

no need to post another link, just read the interview you yourself quoted
slackware won't switch unless PV is forced to and that, if you ask me, won't happen
also that interview was before eudev, afaik

I personally don't care whether Slackware changes to systemd or not, haven't used it since floppies was a way of distributing Linux. The end result seems predictable.

But I think you are misunderstanding Patrick: he likes the way Slackware works, but he doesn't have a suicide pact with people disliking systemd. If Slackware need to change, then he will change it to suit the times, despite him having some reservations.

And since many non-systemd users seems to be in denial of the need to develop or maintain anything, they will not contribute anything towards making it possible for Slackware to avoid switching to systemd.

as for gnome and kde
gnome is the only one now that is dependent
maybe wait for kde to play out then say something

KDE has clearly said many times, just like Gnome, that if non-systemd distros don't help with making it possible for KDE to support them, they can't expect a fully functional desktop.
Bit by bit KDE will start using systemd features (see SDDM), and non-systemd distros doesn't give them any alternative. I don't think you can actually use KDE as a fully functional DE these days on Linux without using systemd code in some form or another.

in the end all that functionality is provided by the kernel (and what isn't is basic logic) so anything can be replaced (even console-kit/logind)

Sure, everything can be replaced, even logind or ConsoleKit. But the fact is that it isn't being done.

nothing in the kernel will ever depend on anything systemd specific, as kernel devs usually have a brain

Yes, the Kernel developers are smart guys, but please notice that several leading kernel developers like Greg Kroah-Hartman (maintainer of the stable Linux kernels, probably the one that takes over Linux kernel development if Linus gets hit by a bus) are working for systemd integration with the kernel etc.

[johnc]

And since many non-systemd users seems to be in denial of the need to develop or maintain anything, they will not contribute anything towards making it possible for Slackware to avoid switching to systemd.

If RedHat was willing to pay me I'd be more than happy to maintain or develop a non-insane alternative to systemd.

[finalzone]

If RedHat was willing to pay me I'd be more than happy to maintain or develop a non-insane alternative to systemd.

Show your skills and proposals. Display your portfolio about your works.

[johnc]

Show your skills and proposals. Display your portfolio about your works.

That's a pretty good point. I should throw something together that's designed to thwart Canonical in some way and I'm sure I'd get an offer from RH the day it hits GitHub.

Not that I would work there anyway.

[gens]

Yes, the Kernel developers are smart guys, but please notice that several leading kernel developers like Greg Kroah-Hartman (maintainer of the stable Linux kernels, probably the one that takes over Linux kernel development if Linus gets hit by a bus) are working for systemd integration with the kernel etc.

to the first part: ye sure whatever
only that kde devs have stated they will accept patches while gnome devs... are gnome devs

as for GKH,
you mean kdbus ?
dbus in the kernel
dbus that systemd uses that was here a long time before systemd ?
that does not change the API ?


anyway

Comparing software engineering to classical engineering assumes that software
has the ability to wear out. Software typically behaves, or it does not. It
either works, or it does not. Software generally does not degrade, abrade,
stretch, twist, or ablate. To treat it as a physical entity, therefore, is
misapplication of our engineering skills. Classical engineering deals with
the characteristics of hardware; software engineering should deal with the
characteristics of *software*, and not with hardware or management.
-- Dan Klein

unless, ofc, the leading devs change the API every once in a while
and that is forbidden in the kernel (for obvious reasons, at least to those with a brain)


fun stuff from https://lwn.net/Articles/580194/
"Unlike most other kernels, Linux has never had a well-designed IPC mechanism."
except NETLINK, but who cares about that (or that dbus is a bit bloated, not that there arnt bloated protocols in the kernel *hint: NFS, but thats from addons*)

[kshade]

Among the expressed objectives of systemd are turning Linux from "a bag of bits into a competitive general purpose operating system", [?] unifying "pointless differences" between distributions [?]

The tasks mentioned that systemd already covers include, "init system, journal logging, login management, device management, temporary and volatile file management, binary format registration, backlight save/restore, rfkill save/restore, bootchart, readahead, encrypted storage setup, EFI/GPT partition discovery, virtual machine/container registration, minimal container management, hostname management, locale management, time management, random seed management, sysctl variable management, and console managment."

And some people wonder why there's concerns about Systemd. They really want distributors to use all their new implementations. This creates an atmosphere of "systemd is standard so let's use all the specific features it has". That in turn creates software that will not or badly run without Systemd. I'm sure this will eventually be solved by creating Systemd-compatible wrappers for "legacy" (Debian and Gentoo are already doing this to some extend) but in the meantime it's really annoying to see project after project jump on the bandwagon or get absordbed.

Seriously, I thought kdbus sounded pretty good until I learned about how userspace programs should interface with it. Of course it's gated by bloody Systemd.

[johnc]

I think at this point X.org is leaner than systemd.

[bkor]

PS gnome devs stated they WONT accept patches to support non-systemd systems, so stop that too

People should just implement an alternative without having loads of #ifdefs all over the place. Someone is actually working on exactly this. So, bzzt, wrong. Kindly don't make up stuff about GNOME, thanks.